Gate2Home · Data tools
Compare two CSV files online by ID or row order. See what changed and download a CSV diff report.
Drop one CSV, TSV or TXT file here.
Drop one CSV, TSV or TXT file here.
UTF-8 CSV or TSV text, not Excel workbooks. Each input: up to 1,000,000 UTF-16 units, 10,000 data records, 100 columns and 100,000 cells including headers. Files: up to 2 MiB. Headers: up to 256 units each. Each complete export: up to 8,000,000 units. Oversized or malformed data is rejected without truncation.
Quoted delimiters, doubled quotes and multiline cells are supported. Blank records are kept and must have the same field count. Delimiters are chosen in Import options; no automatic detection. Comparison runs in this page. Inputs are not saved when you reload.
A CSV diff shows which records were added or removed and which cell values changed. These examples show the inputs and results from this tool. Load one to try the comparison yourself.
The ID stays the same, so these two rows describe the same record. The name is the cell that changed.
| id | name |
|---|---|
| 1 | Tea |
| id | name |
|---|---|
| 1 | Coffee |
1 changed · 0 added · 0 removed · 0 unchanged
“Tea” → “Coffee”Match by SKU to keep each product together when rows and columns move. A different position alone does not count as a changed cell.
| sku | name | price |
|---|---|---|
| A1 | Tea, green | 010.00 |
| B2 | Coffee | 20 |
| C3 | Cocoa | 30 |
| price | sku | name |
|---|---|---|
| 21 | B2 | Coffee |
| 010.00 | A1 | Tea, green |
| 40 | D4 | Water |
1 changed · 1 added · 1 removed · 1 unchanged
“20” → “21”Added in CSV 2: D4.
Removed from CSV 1: C3.
The region column disappears and status is added. A missing column is different from a cell that exists but is empty.
| id | name | region |
|---|---|---|
| 01 | Ada | North |
| 02 | Sam | South |
| id | name | status |
|---|---|---|
| 01 | Ada | active |
| 02 | Sam | empty |
2 changed · 0 added · 0 removed · 0 unchanged
“North” → (column absent)(column absent) → “active”“South” → (column absent)(column absent) → (empty cell)Columns added: status. Columns removed: region.
Use an ID, SKU or another unique key when each row describes an item. The same item stays matched even after sorting the file. If an ID changes, it appears as a removal and an addition.
Use row order when the first data row should match the first data row, and so on. An inserted or moved row can change many pairs. Only extra rows at the end count as added or removed.
With ID matching, rows can be in any order. Columns are matched by their exact header names, so their order can also differ. Added and removed columns are listed separately, and their cells also appear as changed in each matched record. Without headers, columns are compared by their position.
The comparison stops at the first empty, whitespace-only or repeated key and identifies its record or records. Choose additional key columns to form a unique combination, correct the data, or explicitly choose row-order matching. The tool does not guess which duplicate row belongs with another.
No. Values and keys are compared as exact strings. Leading zeros, spaces, capitalization, number spelling and date formatting remain significant. CSV quotes are parsed as syntax, so a quoted value and the same unquoted value compare equally.
Open UTF-8 CSV, TSV or TXT files, or paste their text. Choose comma, semicolon, tab or pipe in Import options for each input. XLS and XLSX workbooks are not supported; export the sheet as CSV UTF-8 first. Quoted delimiters, escaped quotes and multiline cells are supported.
Open “Export the complete report” below the results to download CSV, or copy/download JSON. CSV has one row for each changed, added or removed cell, plus column changes, with source record numbers and matching keys. Fields ending in _json use JSON notation to retain exact text and distinguish empty cells from missing values. JSON keeps the structured groups and summary counts. Exports include all differences within the stated limits, even when the preview shows only the first 20 records.